ROADS AND STREETS. The earliest roads about which any-thing definite is known are those of ancient Rome, one of the oldest of which and the most celebrated for the grandeur of its works - the Appian Way - was com-menced in 312 B.°. Roman roads are remarkable for pre-serving a straight course from point to point regardless of obstacles which might have been easily avoided. They appear to have been often laid out in a line with some pro-minent landmark, and their general straightness is perhaps due to coiwenience in setting them out. In solidity of construction they have never been excelled, and many of them still remain, often forming the foundation of a more modern road, and in some instances constituting the road surface now used. It is consequently possible, with the help of allusions of ancient writers, to follow the mode of construction. Two parallel trenches were first cut to mark the breadth of the road ; loose earth was removed until a solid foundation was reached ; and it was replaced by proper inaterial consolidated. by ramming, or other means were taken to form a solid foundation for the body of the road. This appears as a rule to have been composed of four layers, generally of local materials, though sometimes they were brought from considerable distances. The lowest layer consisted of two or three courses of flat stones, or, when these were not obtainable, of other stones, generally laid in mortar ; the second layer was composed of rubble masonry of smaller stones, or a coarse concrete ; the third of a finer concrete, on which was laid a pavement of poly-gonal blocks of hard stone jointed with the greatest nicety. The four layers are found to be often 3 feet or more in thickness, but the two lowest were dispensed with on rock. The paved part of a great road appears to have been about 16 feet wide, and on either side, and separate(' from it by raised stone causeways, were unpaved side-ways, each of half the width of the paved road. Where, as on many roads, the surface was not paved, it was made of hard concrete, or pebbles or flints set in mortar. Sometimes clay and marl were used instead of mortar, and it would seem that where inferior materials were used the road was made higher above the ground and rounder in cross section. Streets were paved with large polygonal blocks laid as above described and footways with rectangular slabs. Specimens are still to be seen in Rome and Pompeii. There are no traces of Roman influence in the later roads in England, but in France the Roman method appears to have been followed to some extent when new roads were constructed about the beginning of the 18th century. A foundation of stones on the flat was laid, and over that two layers of considerable thickness, of larger and smaller stones, bordered by large stones on edge, which appeared on the surface of the road. In 1764 Tresaguet set the foundation-stones on edge and reduced the thickness of the upper layers, and his method was generally followed until the influence of Macadam began to be felt. A French chaussee with accotements still retains. some resemblance to the old Roman roads.

The almost incredibly bad state of the roads in England towards the latter part of the 17th century appears from') the accounts cited by Macaulay (Hist., c. iii.). It was due 1c chiefly to the state of the law, which compelled each parish to maintain its own roads by statute labour, but the estab-lishment of turnpike trusts and the maintenance of roads by tolls do not appear to have effected any great improve-ment. At the time of Arthur Young's six months' tour in 1770 the roads would seem to have been almost as bad as ever, and it is doubtful if there was much improvement up to the beginning of the present century. The turnpike roads were generally managed by ignorant and incompetent men until Telford and Macadam brought scientific prin-ciples and regular system to their construction and repair. The name of Telford is associated with a pitched foundation, which he did not always use, but which closely resembled that which had been long in use in France, and the name of Macadam often characterizes roads on which all his precepts are disregarded. Both insisted on thorough drainage and on the use of carefully prepared materials, and adopted a uniform cross section of moderate curvature instead of the exaggerated roundness given before ; but, while Telford paid particular attention to a foundation for the broken stone, Macadam disregarded it, contending that the subsoil, however ba(I, would carry any weight if made dry by drainage and kept dry by an impervious covering. Macadam was engaged more with the repair of old roads than ivith the construction of new ones, and, though it is not possible to agree with all his doctrines, the improve-ment which he effected in road management and main-tenance was great and lasting.

Construction, of Roads. - A. road should be as short as possible between two points to be connected, but straightness must often be sacrificed to avoid difficulties a.nd expense and to secure good gradients. The latter should be as easy as practicable, having re-gard to the country to be traversed, and it is desirable that there ( should be a ruling gradient than which none should be steeper. On a level macadamized road in ordinary repair the force which the herse has to put forth to draw a load inay be taken as one-thirtieth of the load. But in going uphill the horse has also to lift the load, and the additional force to be put forth on this account is very nearly equal to the load drawn divided by the rate of gradient. Thus on a gradient of 1 in 30 the force spent in lifting is one-thirtieth of the load, and in ascending a horse has to exert twice the force required to draw the load on a level. In de-scending, on the other hand, on such gradient, the vehicle, when once started, would just move of itself without pressing on the horse. A horse can without difficulty exert twice his usual force for a time, and ean therefore ascend gradients of 1 in 30 on a mac-adamized surface without sensible diminution of speed, and can trot freely down them. These considerations have led to 1 in 30 being generally considered as the ruling gradient to be aimed at on first-class roads, though 1 in 40 has been advocated. Telford adopted 1 in 30 as the ruling gradient on the Holyhead road through North Wales, and there are only two gradients steeper, in places where they were unavoidable. All unnecessary rises and falls should be avoided, but a dead level is unfavourable for drainage, and on this aceonnt 1 in 100 to 1 in 150 is the flattest gradient that is desirable. Such slight rises and falls are probably rather favourable than other-wise to ease of draught by horses.

, In transverse section, roads in tho United Kingdom generally ( comprise the earriage-way, a space on each side, on one or both of s which there may be a footpath, then the fences, and outside all the ditches. The width of the carriage-way may be from 15 feet, which allows of the easy passage of two vehicles, to 30 or 5o feet for roads of irnportance near towns. The side spaces may be from The thickness to be given to a road made altogether of broken stone will depend on the traffic it is intended for. On a good well-drained soil a thickness of 6 inches will make an excellent road fur ordinary traffic, and Macadam's opinion that 10 inches of well-con-solidated material was sufficient to carry the heaviest traffic on any substratum if properly drained has proved to be generally correct. In a new road the loss of thickness during consolidation must be allowed for, and the materials should be laid about one-half thicker than the coating is intended to be. When the materials are not rolled, a thickness of 3 to 6 inches should be laid first, and when that has partly consolidated under the traffic other coats may be added to make up the full thickness. There is great wear and waste of the materials in consolidating if they are laid too thickly at once. Inferior material is sometimes used in the lower part of the road coating, especially when the surface is to be of granite or other hard expensive stone. Thus flints or gravel may be used for the lower Whenever it is possible a new road should be finished with a roller. The materials are consolidated with less waste, and wear and tear of vehicles and horses is saved. Horse-rollers, if heavy enough to be efficient, require a number of horses to draw them and are cumbersome to use. A ton or a ton and a half weight per foot of width is desirable, and to obtain it a roller 4 feet wide must be loaded to 5 or 6 tons, and will require as many horses to draw it. In Great Britain horse-rollers have to a great extent been super-seded by steam road rollers in consequence of the superiority- and economy in the work done. A 15-ton roller, 7 feet wide, giving upwards of 2 tons weight per foot, can thoroughly consolidate 1000 to 2000 square yards of newly-laid materials per day. The materials should be fanned to the proper section, and not more than 4 or 5 inches in thickness ; if a greater thickness is required it is better to roll two coats separately. After several passages of the roller any hollows must be Riled up with small materials, and the rolling must be continued until it causes no motion among the stones. When this result has been attained the binding material may be added. It should be spread dry and uniformly in moderate quantities and should be rolled into the interstices with the aid of watering and sweeping. Provided that all the interstices in the upper stratum of stones are filled after the stones are thoroughly consolidated, the less binding that is used the better. By using binding in larger quantity, and before the stone is thoroughly con-solidated, the amount of rolling required is lessened, but at the expense of durabilify in the road. Watering is necessary from the commencement of the rolling unless the weather is wet, but excessive watering, especially in the earlier stages, tends to soften the foundation.

A pitched foundation like that used by Telford is always desir-able for a road that is subject to heavy traffic. It consists of flat I stones carefully set on edge in courses across the road with the broadest edges downwards. The upper edges should not exceed 4 inches in breadth to hold the broken stone well. All inequalities must be knocked' off, and small stones and chips must be firmly pinned into the interstices with a hammer, so as to form a regular convex surface, with every stone firmly fixed in place. The thick-ness of the pitching is generally 6 or 7 inches; it should not be less than 4, and it may generally be thicker without any sensible increase of cost. At least 4 inches of broken stone are required over the pitched foundation and, when consolidated, 6 are always suffi-cient. A foundation o'f cement concrete 6 inches thick was used by Sir J. Macneill on the Highgate Archway (London) road on a bad clay bottom, and common lime concrete was subsequently used elsewhere. A bed of lias lime concrete 12 inches thick was laid as a foundation in Southwark Street and on the Thames Embankment, Nit it is too expensive for a macadamized road under ordinary cir-cumstances. Burned clay, gravel, or even sand may be usefully employed as a foundation on a clay bottom, to cut off the road material from the clay.

The qualities required in a good road stone are hardness, tough-ness, and ability to resist the action of the weather, and these are not always found together in the same stone. Limestones possess another quality, tbat of furnishing a mortar-like detritus which binds the stone together, and enables it to wear better than a harder material that does not bind. For heavy traffic the best rnaterials are traps, basalts, greenstones, and syenite ; quartzose grits and cherty sandstones are also excellent materials. For moderate traffic the harder limestones are sufficiently durable and make the smoothest and pleasantest roads. Coefficients of quality for various road materials have been obtained by the engineers of the French " Administration des Fonts et Chaussees." The quality was assumed to be in inverse proportion to the quantity consumed on a length of road with the same traffic, and measurements of traffic and wear were systematically made to arrive at correct results. These processes requiring great care and considerable time, direct experiments on resistance to crnshing and to rubbing and collision have also been made on 673 samples of road materials of all kinds. The coefficients obtained by these experiments, which were found to agree fairly well with those arrived at by actual wear in the roads, are summarized in the following table. The coefficient 20 is equiva-lent to "excellent," 10 to "sufficiently good," and 5 to " bad."

Stone for a new road should be evenly broken to a size that will pass every way through a ring 2i inches in diameter. For repairs, especially when the material is tough, a gauge of 2f or 2 inches may be used with advantage, as the stone covers a larger surface, consolidates sooner, and makes a srnoother surface. Stone is best broken by band, but stone-breaking machines have been introduced which supersede hand-breaking to some extent, especially where large quantities of hard stone are to be broken. There is always a certain amount of crushing in breaking by a machine, front which softer stones suffer more, and machine-broken stone is never nearly so cubical, uniform in size, or durable as stone well broken by hand. Broken road material contains about 55 per cent. of solid stone to 45 of void space. In a well-consolidated road the void is filled up by small fragments, detritus, and mud, the result of wear, and specimens of good road surfaces weigh from 93 to 95 per cent. of the weight of the solid stone of which they are made. In the coating. of a well-maintained road the proportion of stones of various sizes varies, but generally from one-third to one-half is found to consist of detritus under three-eighths of an inch in diameter, and there is a very constant proportion of about one-fifth of mud and detritus under one-thirtieth of an inch in diarueter. This appears to be the amount necessary to fill the voids between the fragments of stone when compacted together. In an ill-kept road, from which the mud is not removed, the proportion of detritus is much higher, and mud may constitute nearly one-half of the coating. In pro-portion as the detritus and mud are kept down to the minimum by constant removal from the surface, so will the road be able to resist the action of wet and frost and the wear of the traffic.

The wear of materials, resultiug in their gradual reduction to detritus, is due to the joint action of the traffic and the weather, which cause •surface wear, wear arising from cross breaking, and from rubbing of the stones together. When there is no movement in the body of the road, and the wear is confined to the crushing and griuding at the surface, it is the least possible ; but, when a road is weak from insufficient thickness or solidity on a yielding foundation, bending and cross breaking of the coating take place under passing loads in addition to the surface wear, and the effects are aggravated by the softening action of water finding its way into the road through cracks formed iu the surface and by the disinte-grating action of frost. The wear and waste are thus far larger than on roads of sufficient strength, properly maintained. The destruc-tive effect of wheels is greater as the diameter is less, and to a much greater degyee as the tire is narrower. On hard and strong roads no greater width of wheel than 4i inches is useful, as a wider tire does not bear evenly, but on yielding roads a greater width is of some advantage, though it does not prevent damage from bending and cross breaking of the whole coating under excessive loads. A good deal of attention has been given by French engineers to the measurement of traffic, wear, and the consumption of road materials. Without a knowledge of the anaount of traffic accurate comparisons of wear are-impossible, and an account of the traffic on the roads of France is taken periodically in " collars " or horses drawing loads, and in the Ns-eight drawn. Traffic as measured by weight drawn has of late been observed in some of the streets of London and Liverpool, and has been reduced for comparison to tire weight per foot or yard of width of the carriage-way. Wear may be measured by loss of thickness in the coating ; but the loss of stone in proportion to detritus must also be ascertained before all the effects of wear can be deterniined. The accurate measurement of wear as practised by the French engineers is a complicated process, and it must suffice here to state that 'measured by thickness the wear is seldom found to exceed half an inch, or on the most fre-quented roads of France one inch, of consolidated surface per year, and that about 100 cubic yards of good materials per mile per year are considered as the average consumption under 100 collars of traffic per day. Observations in the United Kingdom on roads well and. systematically maintained have confirmed these results.

The new materials may be added to the road. either in thin coats and small patches year by year or in a thick coat consolidated by rolling. The first method, by which the wear is replaced annually and the traffic is depeuded on to work the materials into the road, can be followed with excellent results, and at no great inconveni-ence to the public under proper management when the traffic is not excessive. Considerable care in the use of materials is required that none may be unnecessarily applied. The annual employment of one-fifth or one-sixth the quantity which it would take to cover the whole surface one stone in thickness is often sufficient to replace wear, and it will then take five or six years to coat every part of the road if it is covered regularly. It is therefore important to apply the new materials only where they are needed, and not to use them where the road is already sufficiently thick. The irregularity of wear and of thickness enables a good roadman to judge where new materials must be applied, and he will apply them in sniall quanti-ties wherever weak places appear. To facilitate this the materials should be placed in heaps by the roadside in the summer, and they should be carefully spread in the autumn and attended to after-wards to ensure consolidation without waste. By good manage-ment a large quantity of materials may be incorporated in a road before the middle of the winter without harassing the traffic, and the strength may not only be maintained but increased. On a hard strong road consolidation may be aided by loosening the sur-face with a pick ; generally only the margin of a patch need be picked up. But if the road is soft or weak it is better not to disturb the surface at all. Binding may sometimes be used to aid con-solidation, but it is seldom necessary if the materials are properly laid and attended to, as the coating already contains detritus enough. In the second method a coating of materials is laid on at once sufficient to endure the wear of several years with such slight repairs as may be necessary to keep a good surface, and, when the wear lias gone as far as it can be safely allowed to go, the process is repeated. Unless the wear is very considerable there is no economy in this method, though the c,onvenience to the public, especially in towns, is undeniable. Consolidation by rolling (after the manner already described) is essential, and it is generally desirable to loosen the old surface to ensure the incorporation of the new coating with it. Scraping and attention are required between one coating and another and also slight repairs to the surface, as, however well the materials may be laid and rolled, the wear of the ordinary timffic will search out places which have escaped the full pressure of the roller and produce inequalities.

Besides a regular application of new materials to replace wear, there must be in road maintenance on proper principles a systematic removal of the detritus by scraping or sweeping, which must be regarded as keeping the whole coating in proper condition, and not as mere surface cleansing. The wear should also be reduced as far as possible by providing sufficient thickness to carry the traffic, by keeping an even surface on which water can never stand and soak, and by good drainage both of surface and subsoil. An adequate amount of skilled manual labour is necessary for economy of main-tenance, and this and the constant attention which is required to keep a road in good order are best secured by putting a man in permanent charge of a defined length. In the autumn and winter, when more labour is wanted, extm men should work under the directions of the permanent road labourer, whose knowledge of his length of road will enable him to employ them to the best advantage.

Concrete macadam, formed by. grouting with lime or cement ( mortar a coat of broken stone lanl over a bed of stone previously a well rolled, has been tried as an improvement on an ordinary r macadamized surface, but not hitherto with much success. When a cleanliness is of importance, and great durability is not required, tar macadam or bituminous concrete may be usefully employed. It is sometimes made by first spreading a coating of broken stone and consolidating it by a roller, and then pouring over it a mixture of coal-tar, pitch, and creasote oil, upon which a layer of smaller stone is spread and rolled in, and the surface finished with stone chippings rolled in. More usually tbe broken stone and bituminous mixture are well incorporated together before they are spread, the stone sometimes being previously heated. The lower layer, about 4 inches thick, may be of stone broken to 21 inches gauge, arid the next layer, about 2 inches thick, may be of smaller stone. Each layer must be well rolled, and when perfectly solid a thin coating of fine stone or granite ehipping,s is spread over the surface and rolled in. Hard limestone is found to be more suitable than sill-clops or igneous rocks for this material. A road surface well made in this manner will last several years under light traffic without any repairs, and it can easily be patched when necessary.

Stone Pavements. - Early pitched roadways consisted of pebbles F, or rounded boulders, bedded in the natural surface or in sand or gravel. The next step in advance was to employ roughly-squared is blocks ; but the wide and irregular joints admitted the water to the subsoil, and the mud worked up and the stones sank irregu-larly under the traffic. Telford, who was called upon to report on the street pavements of the parish of Hanover Square in 1824, saw tke necessity of cutting off all connexion between the subsoil and the paving stones.' He recommended a bed of about 6 inches of clean river ballast, rendered compact by being travelled upon for t some time before the paving was laid, but lie subsequently con-sidered that nothing short of 12 inches of broken stone, put on in layers 4 inches thick and completely consolidated by carriages passing over them, would answer the purpose. He recommended paving stones of considerable depth and of from 4i to 6 or 7A inches in breadth for the greatest thoroughfares, and he pointed. out the importance of working the stones flat on the face and square on all sides, so as to joint close and preserve the bed or base as nearly as possible of the same size as the face, and of carefully placing together in the same course stones of equal breadth. Many pavements thus laid with stones of considerable breadth still re-main, but experience proved that it was a mistake to suppose that broad stones having a larger base would support better the weight and shocks of heavy traffic ; on the contrary, a wide stone has a tendency to rock on its bed, and also to wear round on the top and become slippery. To obtain an evener surface and a better foot-hold for the horses the stones were reduced in width, and in 1840 a granite pavement was laid by Walker on Blackfriars Bridge, which may be considered the first of modern set pavements. The stones were 3 inches broad and 9 deep ; they were laid on a bed of concrete 1 foot thick and were jointed with mortar. The reduc-tion of breadth to about 3 inches was generally followed, but it is only of late years that a concrete foundation has been employed to any great extent, the frequent breaking up to which streets are subject having prevented it. In London a foundation of broken stone has been continued in the chief thoroughfares, the sets being evenly bedded in gravel upon it and rammed with a heavy wooden rammer. Hard core - a mixture of broken stone, clinker, brick rubbish, and old building materials - has also been largely used to form a foundation. In the northern towns of England cinders have been employed, and where the traffic is exceptionally heavy a pitched foundation of stones on edge has been laid when the sets were not paved upon an old macadamized surface. The con-crete for a foundation to a paved street should be made with the best Portland cement, thoroughly mixed in proper proportions with the sand aud gravel or other materials used, water being added as sparingly as possible. A thickness of 6 inches of well-made cement concrete is sufficient for the heaviest traffic, and it can be cut out in slabs for pipe-layinz or repairs and can be relaid and cemented in its place. To obtain the best result a new foundation should not be paved upon for a week. A foundation of bituminous concrete is sometimes used where only a thin bed can be laid, in consequence of there beim,. an old foundation which it is undesirable to disturb. It is maday pouring a composition of coal- tar, pitch, and creasote oil while hot over broken stone levelled and rolled to the proper form, and then spreading a thin layer of smaller broken stone over the surface and rolling it in. It has the advantage that it ciln be paved upon a few hours after it has been laid.

The best materials for pavement sets are the hard igneous and metamorphic rocks, though millstone grit and other hard sedi-mentary rocks of the same nature are used when the traffic is com-paratively light. Excessively hard stone which wears smooth and. slippery is objectionable in spite of its durability. Penmaen-Mawr stone, which is much used in many of the large Lancashire towns, is of this character, and its use was discontinued in London in consequence of its slipperiness and noise. Guernsey granite (syenite) and Mount Sorrel granite (syenite)-have the same nature in a less degree, and in London Aberdeen blue granite is preferred, as, though it wears faster, it keeps a rough surface. Walker's observations on the wear of tram stones showed that Aberdeen granite wore three and a quarter times as fast as Guernsey granite, and in the set pavement of Blackfriars Bridge it was found that after thirteen aud a half years' wear the Aberdeen stone had worn 11 inches, while the Guernsey granite had only worn one-fourth of an inch (equal to •11 and -019 inch per year respectively), or that the former had worn six times as fast as the latter. Observations made by Mr Haywood showed the general rate of wear of Aberdeen granite under heavy traffic in the City of London pavements to have been from '14 to '23 inch per year. The rate of wear of Penmaen-Mawr and Carnarvonshire sets in Liverpool under the greatest traffic is stated to be seldom more than '02 inch per year. A certain proportion between the depth and the length and breadth of sets is required for stability. A shallow stone is more easily tilted up by a heavy weight coming on one edt,,e, and a nar-row stone has a. tendency to turn over sideways. The length, measured across the street, must be sufficient to break joint pro-perly, as two or more joints nerrly in-a line lead to the formation of grooves. For the softer stones a breadth of 4 or 5 inches may be adopted, but for sets of granite or other hard material, with which the joints must be depended on for foothold, the breadth should not much exceed 3 inches. The depth shmild hot be less than twice the breadth, and, as deeper sets weigh more and cost more than shallower ones and the loss by wear is but slight, there is some reason for not exceeding the minimum depth. Where, however, the speedy relaying of a. street pavement is of more im-portance than a saving iu first cost, deeper sets are used, and when they have become so worn as to be uneven the street is relaid with new sets and the old ones are removed to be redressed for use in other streets the sets being used again and again in less irnportant streets as their depth is reduced. In London sets 3 inches wide, 10 to 15 long, and originally 9 deep are used in this manner. In Liverpool sets 4 to 14 inches wide, to average with the joints, 5 to 7 inches long, and 6 to 7i deep according to the traffic am used. In Manchester the sets are 3 to 3i inches wide, 5 to 7 long, and 5 to 6 deep, or 7 in exceptional situations. Sets should lie well squared and not taper from the face downwards ; both joints and face should be free from irregular projections. On a concrete foundation sets are generally bedded on a thin layer of sand or fine gravel ; sometimes they are laid in a bed of fine cement concrete, enough of which is spread over the concrete foundation to be covered while fresh by the sets, which are put in place and smartly tapped, and the joints are grouted at once with cement grout. To allow the cement to become thoroughly set it is desirable that traffic over the paveznent should not be allowed for a fortnight, if that can be arranged. The courses of sets are laid square across the street, no advantage arising from a slanting direction which makes the wear more irregular. At junctions of streets the courses are laid meeting at an angle at the centre line of the narrower street, so that the courses may not run in the direction of the traffic. On steep. inclines the sets are sometimes slightly tilted on their beds, forming a serrated surface to give foothold, and slate has been in-serted in the joints for the same purpose. The water channels are formed by two or three courses of sets laid parallel to the kerb.

Joints simply filled in with gravel are of course pervious to water, and a .grout of lime or cement does not make a permanently water-tight joint, as it becomes disintegrated under the vibration of the traffic. Grouted joints, however, make a good pavement when there is a foundation of concrete or broken stone or hard core. Where there is not a regular foundation imperviousness in the joints is of great importance. In some of the Lancashire towns the joints have for many years past been made by first filling them with clean gravel, well shaken in by ramming, and then pouring in a composition of coal-tar, pitch, and creasote oil. The Manchester pavements are good examples of this system of trusting to impervious jointing to prevent unequal settlement. The founda-tion, where there is not already an old road surface, is a bed of cinders about 1 foot thick, over which are laicl .3 inches of gravel, which are thoroughly consolidated by allowing the traffic to pass over them. The sets are evenly bedded and well rammed after the joints have been filled with clean gravel, ramming and gravel-ling being repeated till the joints are full of gravel. The mixture of coal-tar, pitch, afid creasote oil, well boiled, is then poured over the surface and allowed to percolate and fill up all interstices in the joints, and the pavement is finished by covering it with small gmvel. Joints so formed are imvrvions to wet and have a certain amount of elasticity ; the foundation is kept dry ; and the pavement keeps its form well for many years. The objection is made that in hot weather the composition runs from the joints and makes the streets unpleasant for foot-passengers. This sort of jointing is used in Liverpool and some other large towns, where the sets are laid on a concrete foundation. The elasticity diminishes vibration and noise, and pavements so jointed are said to wear better than others.

A curve like that before described, flattening gradually towards the sides, and having a rise equal to one-sixtieth of the width of , the carriage- way, is a common cross section for a paved street. Sometimes the rise is even less.

A pavement consisting of broad, smooth, well-jointed blocks of granite for the wheel tracks, and pitching between for the horse track, was laid by Walker in Commercial Road (London) for the heavy traffic to the West India Docks in 1825, and similar pavements have been successfully used elsewhere, principally for heavy traffic, in streets only wide enough for one vehicle. In Milan, Turin, and other towns of northern Italy tramways of the same sort are ex-tensively used for the ordinary street traffic. The tractive force required is small, while the foothold on the horse track is good ; but the tram-stones are slippery for horses to pass over. The rigid-ity of the roadway renders it more suitable for slow heavy traffic than for light quick vehicles, and the improvement in other pave-ments has limited the application of this one in ordinary streets.

Wood Paving. - Wood pavements were introduced in England in 1839. Hexagonal blocks of fir, 6 to 8 inches across and 4 to 6 : deep, were bedded in gravel laid on a foundation previously levelled and beaten. The blocks were either bevelled off at the edges or grooved across the face to afford foothold. Other wood pave-ments were tried in London about the same time, but they soou got out of order from unequal settlement of the blocks, aud most of them lasted but a few years. The best of these was Carey's, which consisted of blocks 61 to 7i inches wide, 13 to 15 long, and 8 or 9 deep, the sides and ends having projecting and re-entering angles locking the blocks together with the view of preventing unequal settlement. Pavements on this system were laid in Minc-ing Lane in 1841 and in Gracechurch Street in 1842. In the latter street the blocks appear to have been relaid every three or four years and to have been entirely removed about every eleven years, until the pavement was removed in 1871, to be sliperseded. by asphalt. Experience led to a reduction in the width of the blocks to 4 inches and in the depth to 5 or 6, and the salient and re-entering angles disappeared from the sides. With these modifications Carey's pavement remained in use from 1841 until after the introduction of more modern systems in recent years. The " improved wood pavement " was first used in London in 1871. After the foundation was formed to the proper cross section a bed of sand 4 inches deep was laid, upon which came two layers of inch deal boards saturated with boiling tar, one layer across the other. The wooden blocks were 3 inches wide, 5 deep, and 9 long ; they were dipped in tar and laid on the boards with the ends close together, but transversely the courses were spaced by fillets of wood three-fourths of an inch wide nailed to the floor and to the blocks. The joints were filled up with clean pebbles' rammed in and were run with a composition of pitch and tar, the surface being dressed with boiling tar and strewed with small sharp gravel and sand. In this pavement a 'somewhat elastic foundation was pro-vided in the boards, which were also intended to prevent unequal settlement of the blocks; but the solidity of the pavement depended upon its water-tightness, for, when the surface water reached the sand, as it clid sooner or later, settlement and dislocation of the blocks under the traffic arose. Pavements on this system were laid between 1872 and 1876, and were kept in repair and relaid from time to time, but about 1877 the plank foundation was abandoned for a foundation of cement concrete.

A concrete foundation for a wood pavement appears to have been ] first employed in a pavement laid in 1872 in Gracechurch Street by i the Ligno-Mineral Company. The concrete was of blue lias lime 4 inches thick formed to the curve of the road. The blocks were of beech, mineralized by a special process, 3-i inches wide, 4i deep, and 7A long, with the ends cut to an angle of 60°, so that each block might derive support from the next one. They were laid with the ends inclining in opposite directions in alternate courses. The upper edges of the blocks were chainfered, and there was a cham-fered groove near the bottom. In a few years this form of block was abandoned for rectangular blocks, and mineralized fir was sub-stituted for beech. The blocks were bedded in Portland cement and laid with joints one-fourth of an inch wide, partly filled with asphalt and then grouted with mortar. The adoption of a bed of concrete as the weight-bearing foundation of the road marks a new departure, and in all the more recent systems of wood pave-ment a substantial foundation of concrete is an essential feature. In .Norwich, however, a large quantity- of wood paveinent has been laid on the old street foundation, the blocks being bedded in gravel and s.and and rammed, and the joints grouted Nntli lime and sand. The experience of from four to seven years has proved the pavements to be successful, but the foundation is exceptionally dry and hard and the traffic not very heavy. With a concrete foundation there is no reason for complicated shapes and contrivances for locking the blocks together ; and wood pavements in their modern form consist of rectangular blocks (obtained by cutting off tbe end of a deal plank), bedded on the concrete with the fibres of the wood vertical, thus constituting a slightly elastic wearing surface on a rigid foundation, by which the weight of the tmffic is borne. There is, however, considerable variation in the method of bedding and jointing the blocks. The Asphaltic Wood Pavement Company laid half an inch of a.sphalt upon the concrete, and formed the lower part of the joint of asphalt and the upper part of a grout of Port-land cement and gravel, the advantage claimed behie- a slightly elastic bed for the blocks and water-tight joints. The'blocks have been laid in unset cement over the concrete and rammed to au even surface ; but tbe ramming is liable to split the block-s, and the indentations formed in the cement surface of the foundation have to be removed when the time comes for renewing the blocks. It is now more usnal to bed the blocks directly on the concrete, a smooth surface being formed either with the concrete itself or by a floating of cement, and to fill the joints with a grout of cement and gravel. A cement joint adheres to tbe blocks, resists wet, and does not we,ar down too much below the surface of the wood, anti so form a receptacle for mud. In Ilenson's systern, which has been largely used, the blocks are bedded and jointed with ordinary roofing felt, a strip of which, cut to a width equal to the depth of the blocks, is placed between every two courses. Thejoint is made as close as possible by driving up the blocks as evely eight or ten courses are laid with heavy mallets, - a plank being laid along the face of the work. A perfectly close and slightly elastic joint is thus formed. A continuous layer of felt is likewise laid over the concrete foundation to give a slightly elastic bed to the blocks. A V-shaped groove along the centre of every fourth block was at first considered necessary for foothold, but its use has been dis-continued except on gradients steeper than 1 in 30. The surface of the pavement is dressed over with a hot bituminous compound, and covered. with fine clean grit. This method of laying a wood pavement, although somewhat more expensive, is probably the best that has hitherto been devised for smoothness and durability. The blocks are laid in courses across the streets, any change in the direction of the latter being accommodated by shorter courses ending with wedge-shaped blocks. At street junctions the courses are laid diagonally, or meeting at right angles. Two or three courses are laid parallel with the kerb to form a water channel. The blocks may be laid close end to end across the street if sorne allowance be made for expansion by wet, without which the kerb-stones and footways will be displaced, or the courses will be bent in reversed curves. To afford relief the joints of the courses parallel to the kerb may be left open, or the course next the kerb may- be left out until expansion has ceased, the spaee being temporarily filled in with sand. In the direction of the traffic Joints more or Iess wide are generally thought necessary for foothold. A wide joint allows the fibres of the wood to spread and give way at the upper corner of the blocks for want of lateral support, and it also forms a receptacle for mud and wet Experience has shown that the space of three-fourths of an inch or one inch, once thought necessary for foothold, may safely be reduced to one-fourth or three-eighths of an inch. For spacina the courses to form the joints strips of wood of the proper thitkness may be laid in and removed before the joints are filled, or they may be nailed to the lower part of the blocks. Two fillets have been nailed on, or three cast-iron studs fixed in the sides of each block to keep them steady in place until the joints are filled and thoroughly set. The latter method secures more uniformity in the width of the joints. f. There is some difference of opinion as to the best material for a wood pavement. Pitch pine and the harder red and yellow deals are the most durable, but they are less elastic than the softer woods, and are apt to rear slippery. Soft white woods have been recom-mended for the sake of a more elastic surface ; but on the whole either Memel or Swedish yellow deal is generally considered the'best material. Whatever wood is used, it should be sound, close-grained, even in quality, free from knots and sap, and from the blue tinge which is a sign of incipient decay. After the blocks are cut, all those that aro unsound, knotty, or badly shaped should be care-fully rejected, as defective blocks soon cause holes in the surface and must be replaced, or the adjoining blocks will suffer undue wear and the surface become irregular. The breadth of the blocks never now exceeds 4 inches, and It is generally 3, the length being determined by the breadth of the deal or batten from -which they are cut. The depth is usually 5 or 6 inches ; 5 inches are con-sidered by many to be enough to give sufficient depth for as long as the pavement will retain a sufficiently good surface without renewing the wood, and ljlocks of that depth have been laid in many London streets. It is doubtful if any advantage is derived from creasoting or front dipping the blocks in ereasote oil or coal tar. Dipping affords a cover for the use of defective or inferior wood, and thorough ereasoting, though it preserves the wood from decay, has little or no influence on the wear, which in almost all eases d.etermines the life of the blocks.

With a curved cross section like that already described a rise C from the mean level of the channels to the crown of the road equal s, to one-sixtieth or one-seventieth of the width of the carriage-way is enough. The necessary profile must be accurately given to the concrete foundation when wet. Wooden moulds or templates are fixed across the street 10 or 12 feet apart, over which a straight batten is worked to give the concrete the required forra and a smooth surface. The moulds are removed when the concrete is partially set and the spaces are made good with cement mortar. In a level street provision should be made in the foundation for a I fall in the side channels towards the gullies of not less than 1 in t 150, and the necessary modifications of cross section at the inter-section of streets must also be provided for. Every care should be taken to ensure a good homogeneous concrete for the foundation, as upon that the strength of the road depends. With a well-made Portland cement concrete a thickness of 6 inches is sufficient. It should be allowed to set thoroughly before the blocks are laid, and traffic should not be allowed to pass over it for a week. The finished pavement should be covered with a thin layer of sharp grit, which is forced into the wood by the tmffic and forms a hard face. Several applications of grit are desirable at first, and from time to thne afterwards, both as a protection to the wood and to prevent slipperiness. Systematic cleansing is required to prevent slipperiness and foul smells, and to preserve the pavement. Cleans-ing may be aided by washing, and when it is thoroughly canied out but little watering is required to keep down the dust. A wood pavement is the quietest for the residents, pleasant to travel over, and favourable to the wear of vehiles. Traction on it is easy and foothold good, so that it may be laid on gradients as steep as 1 in 20.

The wear of wood pavements in London is stated by Mr Stayton to be from -065 inch per year in Sloane Street, with a traffic of 279 tons per yard„of width per day, to .456 inch per year in Fleet Street, with a traffic of 1360 tons per yard of width per day. Reduced to a standard of traffic of 750 tons per yard per day, the comparative annual wear becomes -175 in the former and -251 in the latter street. In Parliament Street, Westminster, blocks removed after four years in places where patching was required had lost I+ to 4 inches in thickness, equal to one-third of an inch per year under traffic stated to be 1106 tons per yard of width per day. From information afforded by Mr Haywood it appears that in the City of London under traffic of from 300 to 660 vehicles per y-ard of width per day of 12 hours the wear is from -2 to -3 inch per year, and that in King William Street, London Bridge, under a traffic of about 1200 vehicles per yard of width in 12 hours the wear was found to be 2i inches in 3* years in the middle of the road, or -81 inch per year. This is the heaviest traffic to which wood pavement has been sub-jected. The wear is generally considered to be as much due to the horses' feet as to the wheels, and the action of the former is more destructive on steep gradients. Towards the end of the life of the blocks tbe wear is more rapid than at first. Few wood pavements retain a sufficiently good surface after about six years' rear without extensive repairs, and it is probably not advantageous to lay blocks of a greater depth than will provide for a duration of seven years ; 5 inches are almost always sufficient for this.

Wood pavements of plain blocks on a cement concrete bed are ( now (1885) laid at from 10s. 6(1. to 12s. 6d. per square yard, a con-siderable reduction on the prices paid for patented systems a few years ago. Of the above prices 2s. 3d. to 3s. 9d. is the cost of the foundation, which does not require renewal like the blocks. As-suming the average life of the latter to be seven years, Mr Stayton estimates the annual cost of wood paving in Chelsea vith a traffic of 500 to 750 tons per yard of width per day to bc Is. 9d. per square yard, which includes the cost of original construction, repairs and renewals, and interest, spread over fifteen years. Cleansing and sanding are estimated to cost 5d. per square yard in addition.

Asphalt Paving. - Asphalt was first used for street paving in Paris in 1854. It was introduced in London in 1869, when Thread- r needle Street was paved by the Val de Travers Asphalt Cornpany„ and since then it has been extensively used for paving both streets and footways. The material is a bard limestone impregnated with bitumen in the proportion of from 6 to 8 per cent. in the Seyssel rock, and from 10 to 12 in that from Val de Travers. Asphalts con-taining less than the former proportion have not sufficient colier-C11Ce for street pavements, and those containing more than the latter proportion soften from heat in the summer. Asphalt is empio) Comparison cf Street Surfaces. - The comparative cost of various street surfaces in Liverpool, including interest on first cost, sinking fund, maintenance, and scavenging, when reduced to a uniform standard traffic of 100,000 tons per annum for each yard in width of the carriage-way, is given by Mr Deacon as follows : - Per square yard per year. Set pavement of bard granites „ „ softer granites ls. 2td.

Bituminous concrete ls.

Wood pavement 2s. 21d.

Macadam, on hand-pitched foundation 2s. Hid.

Taking the standard of traffic at 40,000 tons per annum for each yard in width, the cost for the last three pavements is : - Per sqnare yard per year.

Bituminous concrete ls.

Wood pavement ls.

Macadam Is. Did.

Asphalt pavin a may be placed between wood and bituminous con-crete in the abOve order. These comparisons show the high cost of a macadamized suwiace in a street where the traffic is great. However well it may be maintained, a macadamized street must be dirtier and dustier than any pavement, though it is superior to them all in safety and to set pavements in the matter of noise. Bituminous concrete or asphalt macadam is cheaper, cleaner, and quieter than ordindry macadam and is sufficiently durable when the traffic is not heavy. For heavy traffic no pavement is so econo-mical as granite sets ; but for the sake of quiet and cleanliness a wood or asphalt pavement is often preferable. Asphalt can be kept cleaner than any other pavement and is the pleasantest to travel over ; wood, on the other band, is quieter for the residents, less slippery, and can be laid on steeper gradients.

The comparative ease of draught on various surfaces is largely ] influenced by the amount of foothold afforded, and it may be doubted if dynamometer experiments, however carefully made, are altogether conclusive. The tractive force is influenced by the g,radient, the diameter of the wheels, the friction of the wheel axles, and the speed, as well as by the resistance of the road surface, and these must be all taken into account to obtain accurate results. Some recent expeiiments made, under the direction of Sir J. W. Bazalgette, with Easton and Anderson's horse dynamometer on London street surfaces gave the following mean results : - Tractive force on the level. Macadainized surface 401 to 44'29 lb per ton.

Asphalt 39'0 „ 39*32 „ Wood 33•62 „ 36•63 „ Granite sets 2(5'2 „ „ The gross load was 4 tons, drawn at a speed of fi•om 2 to 6 miles an hour. It is remarkable that the tractive force on asphalt is so high • but the other results are consistent with former experiments by dorin, Macneill and others: The comparative safety of granite, ivood, and asphalt pavements: in the City of London was the subject of careful observations, which were fully reported on by Mr W. Haywood in 1873. The pavements selected were granite sets 3 inches wide, ligno-rnineral pavement of beech blocks 3i inches wide, improved wood paveinent of fir blocks 3 inches wide, and Val de Travers compressed asphalt pavement. On known lengths of these the traffic, the accidents to horses, the weather, and other circumstances were observed for fifty days, and when the distance traversed was taken into account it was found that as a mean result a horse might be expected to travel 132 miles on granite without fallin„,0 191 on asphalt, and 446 on the im-proved wood pavement. The condition of the weather had consider-able effect : on the granite when dry a horse might be expected to travel 78 miles without falling, when damp 168, and when wet 537 ; on wood when damp 193 miles, when wet 432, and when dry 646 ; on asphalt when damp 125 miles, when wet 192, and when dry 223. It thus appeared that wood pavement was less slippery than either g,ranite or asphalt in a marked degreo, it being only more slippery than granite when both pavements ivere wet. About 85 per cent. of the falls on the wood pavement were falls on the knees, which are less likely to injure the horses and are less incon-venient to the traffic than other falls. On the granite the falls were falls on the knees or complete falls in about equal propottions, with about 7 per cent. of falls on the haunches. On the asphalt 43 per cent. were complete falls and 24 per cent. falls on the haunches.

Watering. - On macadamized roads in Great Britain watering is only good for the road itself when the materials are of a very sili-eious nature and in dry iveather. With other materials the effect is to soften the road and increase wear. In and near towns water-ing is required for the comfort of the inhabitants, but it should not be more than enough to lay the dust without softening the road, and the amount required for this may be greatly reduced by keeping the surface free from mud, and by sweeping off the dust when slightly wetted. Pavements are watered to cleanse them as well as to lay the dust, but it must be remembered that both wood and asphalt are more aippery when wet, and that therefore watering should be obviated as far as possible by thorough cleans-ing. Hydrostatic vans, by improvements in the distributing pipes and regulating valves, water a wide track uniformly with an amount of water which can be regulated at pleasure. Where hydrants exist in connexion with a water supply at hig•h press-are, street watering can be effected by a movable hose and jet, a method much more effective iu cleansing the surface, but using a much larger quantity of water. Another method which has been tried, but not much used, is to lay perforated pipes at the back of the kerb on each side of the road, from which jets are throivn upon the surface. The first cost is considerable, and the openings for the jets are liable to choke and get out of order. Deliquescent salts have been used for street watering, by which the surface is kept moist, but at the expense of the moisture in the air. Sea water has the same effect in a less degree.

Cleansing. - The principal streets of a town Etre generally cleapsed daily, either by hand-sweeping and hand-scraping or by machines. Whitworth's machine consists of a series of revolving brooffis .on an endless chain, whereby the inud or dust is swept up an ffichne into the cart. A less costly and ctunbersome machine consists of a revolving brush mounted obliquely, which sweeps a track 6 feet wide and leaves the dust or mud on one sib to be gathered up by hand. A horse scraping-machine which delivers the mud at the side is also used, the blades of the scrapers being mounted obliquely and covering a width of 6 feet, For general use, more especially in the country, scraping machines worked by a man from side to side of the road, and scraping a width of about 4 feet, are more convenient.

All street surfaces stiffer from the constant breaking up and dis-turbance to which they are subjected for the purpose of laying and repairing sus and water pipes. Subways, either under the middle of the roa.d or near the kerbs, in which the pipes may be laid and be always accessible, have often been advocated, and in a few instances have boen constructed ; but they have not hitherto found general favour.

Footways. - Grav,e1 is the most suitable material for county or suburban footways ; it should be bottomed with a coarser material, well drained, and should be laid with a. roller. An inclination towards the kerb of about half an inch in a foot may be given, or the surface may be rounded, to throw off the wet. Where greater cleanliness is desirable and the traffic is not too great a coal-tar concrete similar to that already described, but of smaller materials, makes a good and economical footway. The coating should be 2i or 3 inches thick, composed of two or three layers each well rolled, the lower layer of materials of about inches gauge, and the upper of a half or a quarter of an inch gauge, with Derbyshire spar, or tine granite chipping,s over all. Concrete footways require to be care-fully made and must be allowed to set thoroughly before they are used. Concrete has a tendency to crack from contraction, especially when in a thin layer, and it is better to lay a footway in sections, with joints at intervals of about 2 yards. Concrete slabs, especially when silicated and constituting artificial stone, make an excellent footway. The material is composed of crushed granite, gravel, or other suitable material, mixed with Portland cement and cast in inonlds, and when set saturated with silicate of soda. This paving has proved more durable than York stone flagging, but it is more slippery, especially when made with granite. York stone makes a good and pleasant foot pavement, but is somewhat expensive con-sidering its durability ; it is apt to wear unevenly and to scale off when the stone is not of the hest quality. It should not be laid of a less thickness than 2 inches ; 2i or 3 inches are more usual. The flags should be square jointed, not under-cut at the edges, and should be well bedded and jointed with mortar. Caithness flag is much more durable than York stone and wears more evenly ; it is impervious to wet and dries quickly by evaporation. The edges are sawn, and the hardness of the stone renders it difficult to cut it to irregular shapes or to fit openings. Staffordshire blue bricks and bricks made of scoria from iron furnaces are both very durable, though somewttat brittle. Asphalt either laid as mastic or com-pressed is extensively nsed for footways ; the former is considered inferior in durability to York stone and the latter superior to IL Asphalt should not be laid less than three-fourths of an inch thick on 4 inches of cement concrete, and 1 inch of asphalt is desirable where there is great traffic.

Footways in a street must be retained by a kerbing of granite, York stone, Purbeck, or other stone sufficiently strong to stand the blows from wheels to which it is subjected. It should be at least 4 inches wide and 9 deep and in lengths of not less than 3 feet. A granite kerb is usually about 12 by 6 inches, either placed on edge or laid on the flat. When set on edge a kerb is generally bedded on gravel with a, mall; when laid on the fiat a concrete bed is desirable.

In a macadamized street pitched or paved water channels are required, to prevent the wash of the surface water from under-mining the kerb. The pitching consists of cubical blocks of hard stone about 4 inches deep, bedded on sand or mortar, or preferably on a bed of concrete. A paved channel consists of fiat stones about 1 foot wide inclining slightly towards the kerb. Moulded bricks and artificial stone are also used both for side channelling and for kerbing. Such an inclination must be given to the channel as will bring the surface water to gullies placed at proper intervals, and the level of the kerbing and consequently of the footway will depend • to some extent on the surface drainage as well as on the levels of adjacent houses. To lay out a street satisfactorily the longitudinal and transverse sections must be considered in relation to these matters as well as to the levels of intersecting streets.

For fuller information on the subject see Sir IIenry Parnell, A Treatise on Roads; Thomas Codrington, 77te Maintenance of Macadamized Roads; Debanve, Manuel de l'Inelnieur des Ponta et Chausses ; Ansudes des Pouts et Chaussies; Minutes of Proc. Inst. Civ. Eng., "Street Pavements," vol. WM, p. 1, and "Wood Pavements," voL lxxviii. p. 240; Reports by w. Haywood, engineer to the commissioners of sewers of the City of London. (T. C.)